2.23) where a is the distance between the spins, A is called the exchange stiffness constant having
2.4. Anisotropy
When a physical property of any material depends on the direction of the materials that
property is said to exhibit anisotropy. In magnetism, the preference of magnetization to lie
in a particular direction of the sample is called as magnetic anisotropy. As anisotropy plays
an important role in tuning the nature of MβH loop, it is much essential to understand the
various possible sources of the magnetic anisotropy and its influence on the control of the
magnetic properties. Figure 2.06 displays a typical situation where for zero applied field,
the magnetization would point along the easy axis shown (ο‘ = 0). When a field is applied,
the magnetization is pulled towards the field direction and approaches closer to the field direction with increasing the applied field. For any intermediate values of
ο‘, themagnetization is being attracted in opposite directions, i.e., up by the field and down by the anisotropy.
Let us assume that all the magnetization is pointing in the same direction in a magnetic material and the material exhibits an easy axis of magnetization. In such scenario, we can describe the energy per unit volume of the magnetization of this material by
πΈ = πΎ π ππ2πΌ
(2.24)
where K is called anisotropy constant with a unit of energy per unit volume (J/m
3or ergs/cc).
Hence, the energy term,
E, is also energy per unit volume. In general, the magnitude ofuniaxial anisotropy is described in terms of the anisotropy field, which is defined as the field needed to saturate the magnetization of a uniaxial crystal in the hard axis direction, as given in eqn.(2.25)
π»π = 2πΎπ
π0π
(2.25)
In general, the energy of the magnetization is given by,
πΈ = πΎ π ππ2πΌ β π0ππ» πππ (π½ β πΌ)
(2.26) where the first term is anisotropy energy. The second term is due to the magnetic field and the difference in the angle (ο’ - ο‘) is the angle between H and M. In order to get equilibrium, the first derivative is required to be zero.
Figure 2.07: Magnetization of single crystals of iron, nickel and cobalt [HTTP2].
Therefore, taking derivative of eqn.(2.26) with respect to the angle provides,
ππΈ
ππΌ = 2πΎ sin πΌ cos πΌ β π0ππ» π ππ(π½ β πΌ) = 0
(2.27) Taking the value of ο’ as 90
ofor the equilibrium angle for the magnetization relative to the easy axis and considering the eqn.(2.25) givesο¨
sin πΌ = π»
π»π
(2.28)
The above eqn. indicates that when field is zero, the magnetization points along the easy axis and when the field is equal to H
k, the magnetization points along the direction of field.
For any intermediate value of the applied field, the magnetization points at a value of angle given by eqn.(2.28) rotating smoothly between the easy axis and the applied field.
Figure 2.08: Schematic drawing of broadside and head-to-tail configurations for a pair of ferromagnetically coupled magnetic moments.
2.4.1. Magnetocrystalline anisotropy
Figure 2.07 depicts initial magnetization curves of single crystals of different 3d ferromagnetic elements. It is seen that the materials approach to saturation in different ways when magnetized in different directions. For example, iron display a ο‘100ο± as easy directions and
ο‘111ο± as hard directions, while nickel exhibits ο‘111ο± as easy axis and ο‘100ο± as harddirections. This behaviour can be understood by analyzing the development of anisotropy energy in different symmetries as given below:
For Hexagonal:
πΈπ = πΎ1sin2π + πΎ2sin4π + πΎ3sin6π + πΎ3β²sin6π sin 6π
(2.29) For Tetragonal:
πΈπ = πΎ1sin2π + πΎ2sin4π + πΎ2β²sin4π cos 4π + πΎ3sin6π
+ πΎ3β²sin6π sin 6π
(2.30)
For Cubic:
πΈπ = πΎ1π(πΌ12πΌ22+ πΌ22πΌ32+ πΌ32πΌ12) + πΎ2π(πΌ12πΌ22πΌ32)
(2.31)
where
ο‘iare the direction cosines of the magnetization.
K1cterm is equivalent to
πΎ1π(sin2π πππ 2π sin2π + cos2π sin2π). When, ο±= 0,
ο¦= 0, this term reduces to eqn.(2.24) [COEY2010].
Origin of magnetocrystalline anisotropy: There are two distinct sources of
magnetocrystalline anisotropy: (i) single-ion contributions and (ii) two-ion contributions.
The first one is essentially due to the electrostatic interaction of the orbitals containing the magnetic electrons with the potential created at the atomic site by the rest of the crystal.
This crystal field interaction stabilizes a particular orbital and by spin-orbit interaction, the magnetic moment is aligned in a particular crystallographic direction. For example, a uniaxial crystal having 2ο΄10
28ions/m
3described by a spin Hamiltonian DS
2with D/k
B= 1 K and S = 2 will have anisotropy constant K
1= nDS
2= 1.1ο΄10
6J/m
3. On the other hand, the later contribution replicates the anisotropy of the dipole-dipole interaction. Considering the broadside and head-to-tail configurations of two dipoles each with moment m, as shown in Figure 2.08, the energy of the head-to-tail configuration is lower by
3π0π2/(4ππ3) andhence the magnets tend to align head-to-tail. In non-cubic lattices, the dipole interaction is an appreciable source of ferromagnetic anisotropy.
Figure 2.09: Magnetization of a prolate ellipsoid of revolution with
c > a and nomagnetocrystalline anisotropy. c-axis is the easy direction of magnetization.
2.4.2. Shape anisotropy
Shape anisotropy arising due to the asymmetric shape of the material is important for thin
films where one dimension is limited as compared to other two dimensions. The
demagnetization field inside the material or the stray field outside the magnetic material
depends on the magnetization and shape of the material [JILE1997, OHAN2000,
BLUN2001]. The magnetostatic energy of a ferromagnetic ellipsoid (see Figure 2.09) with
magnetization M
Sis given as
πΈπ =1
2π0ππππ2
(2.32)
The anisotropy energy is related to the difference in energy
οE when the ellipsoid ismagnetized along its hard and easy directions. N is the demagnetization factor tensor for the easy direction. Nο’=(1/2)(1-N) is the demagnetization factor tensor for the hard directions.
Hence,
βπΈπ =1
2π0πππ2[1
2(1 β π) β π]
βπΈπ = 1
4π0πππ2[1 β 3π]
πΎπ β = 1
4π0ππ2[1 β 3π]
(2.33)
Table 2.01: Demagnetization factors (in Gaussian units) of selected shapes:
Shape
N1 N2 N3Sphere 4ο°/3 4ο°/3 4ο°/3
Long Cylinder along z-axis 2ο° 2ο° 0
Infinite plate normal to z-axis 0 0 4ο°
Strip film normal to z-axis
(with t β thickness, W β Width, L β Length; t ο£ W ο£ L)
0 8t/W 4ο°
In addition, the demagnetization factor tensor that relates the demagnetization field with a specimen magnetization as a function of position is given by [NEAL1994]
π(π) = β 1
4πβ π3πβ²ββ² (ββ² ( 1
π β πβ²))
(2.34)
This tensor is given by an integral over the object volume and can be evaluated either inside or exterior to the body. The value of tensor N significantly depends on the specimen shape, which is difficult to obtain in closed-form. It may be calculated exactly for an ellipsoidal shape only. In many symmetrical materials such as any ellipsoid of revolution, the demagnetization factor tensor only has three principal components, i.e.,
( π»1 π»2
π»3) = β (
π1 0 0 0 π2 0 0 0 π3) (
π1 π2
π3)
(2.35)
where
N1+ N
2+ N
3= 1 (in SI) and
N1+ N
2+ N
3= 4ο° (Gaussian). The demagnetization factors for the selected shapes are summarized in Table 2.01. A detailed calculation of demagnetization factor for various objects can be found in [NEAL1994]. The shape anisotropy is zero for a sphere, as
N = 1/3. Shape anisotropy is fully effective in sampleswhich are so small that they do not break up into domains [COEY2010].
Figure 2.10: Magnetization of a thin film with induced anisotropy created by annealing in a magnetic field. The sheared (open) loop in a (b) is observed when the measuring field H is applied perpendicular (parallel) to the annealing field direction.
2.4.3. Induced anisotropy
In some materials, the magnetic anisotropy can be induced by many ways: (i) fabricate a film in the presence of a magnetic field, (ii) post annealing the materials in the presence of magnetic field and (iii) apply uniaxial stress. In the first two cases, after such treatment, the material may exhibit an easy axis of magnetization that points in the direction of the magnetic field. This induced anisotropy is certainly independent of any crystalline anisotropy or any other form of anisotropy that might be present. Figure 2.10 shows the typical example of inducing the anisotropy in ferromagnetic materials by field annealing.
In the last case, the uniaxial anisotropy is induced by applying uniaxial stress (ο³) in a ferromagnetic solid [KRON2003]. The magnitude of the stress-induced anisotropy is
πΎπ’π =3
2πππ
(2.36)
where
ο¬Sis the saturation magnetostriction. Both the single-ion and two-ion anisotropy
contribute to the stress induced anisotropy. The highest values of uniaxial anisotropy are
found in hexagonal and other uniaxial crystals. Smallest values are found in cubic alloys and amorphous ferromagnets.
Figure 2.11: Schematic drawing of bars to demonstrate inducing an easy-axis in a material with the positive magnetostriction.
2.4.4. Magnetostrictive anisotropy
Another important form of anisotropy in magnetic materials is due to magnetostriction, a change of volume of an isotropic crystal due to magnetic order. Magnetostriction relates the stress in a magnetic material to an anisotropy created by that stress. Figure 2.11 shows the schematic views of bars with different applied stress conditions. If
ο¬ is positive, thenapplication of a tensile stress to the bar creates an easy axis in the direction of the applied stress. If a compressive stress is applied, then the direction of the easy axis created will be perpendicular to the stress direction. On the other hand, if the magnetostriction constant for the material is negative, then the above phenomena would be reversed: a tensile stress will create an easy axis perpendicular to the stress direction and a compressive stress will create an easy axis in the direction of the applied stress.
2.4.5. Magnetic surface anisotropy